The proposed work is part of a continuing effort to understand how integrative tasks are performed in the mammalian cochlear nuclear complex. We hope to contribute to an understanding of what strategies the brain uses to process acoustic information by examining how neurons are interconnected. In the coming grant period two goals will be pursued. First, experiments will be performed to examine the pattern of convergence and divergence. On the basis of previous work which showed that anatomical cell types can be identified by their physiology and which revealed the location of local collaterals, we will inject pairs of cells that are likely to be connected. Estimates will be made of the proportion of terminals of one cell that are apposed to labeled processes of the other at the light microscopic level. We will also combine the labeling of single cochlear nuclear neurons intracellularly with intraaxonal or small extracellular injections to examine the innervation of cochlear nuclear neurons by auditory nerve fibers. With these experiments we will visualize connections between cells and relate them to synaptic responses recorded in these and in previous studies. Second, we will focus on the physiology of granule cells. These cells are numerous and occupy prominent positions in the cochlear nuclei of most mammals, yet it is not known how they are driven. The small size of granule cells makes electrophysiological studies difficult. We will attempt to measure extracellularly their spontaneous activity and their synaptic responses to shocks of various parts of the slice. We will also develop patch-clamping techniques to record from them intracellularly.